Carotid imaging after TIA or minor stroke
NICE NG128Imaging exists to answer one question — is there a surgically treatable stenosis on the symptomatic side — and to answer it inside days, because the benefit of endarterectomy falls away rapidly after the index event. That deadline is what makes the most accessible adequate test the right one.
A carotid territory transient ischaemic attack or non-disabling ischaemic stroke in a patient who would be a candidate for carotid intervention.
Referenced decision support — confirm against your local protocol.
Decision support, not a directive. Protocols and timings shown are typical published ones — your local protocol takes precedence, and the vetting radiologist decides.
Accept as requested
Carotid duplex ultrasound
Carotid Doppler ultrasound
What we'd amend, and why
- Duplex infers the degree of narrowing from what the blood does rather than from what the lumen looks like: flow through a fixed narrowing must accelerate to preserve volume, so peak systolic velocity in the internal carotid rises predictably with stenosis, and the ratio to the common carotid corrects for the patient's own cardiac output. It needs no contrast and no radiation, can be done at the bedside within hours, and it also reports plaque surface morphology and the direction of vertebral flow. Its weaknesses are physical: dense shadowing calcification hides the very segment being measured, and velocity thresholds vary between laboratories, which is why the report must state the criteria used.
Nothing needs resolving before this goes ahead. Routine checks below are owned downstream.
How was this decided?
- pathwayadult — Adults
Decision support only. Local protocol takes precedence.
Pathways
Big forks are separate pathways; the first whose conditions match is the one used.
Within a day of onset — still inside the hyperacute stroke pathway
| Role | Study & protocol | Why this answers the question |
|---|---|---|
| First line | CT Angiogram — Carotid and Vertebral (Arch to Vertex) CT Angiogram — Aortic Arch to Vertex usually appropriate | Inside the first day the deciding factor is access, not elegance. The benefit of endarterectomy is concentrated in the days immediately after the event and decays steeply thereafter, so the study that answers the question today beats the study that answers it more tidily next week — and in a patient who presented hours ago, the arch-to-vertex angiogram is frequently already being acquired as part of the hyperacute stroke assessment, which makes the carotid answer a matter of extending an acquisition rather than booking an appointment. It also covers what duplex cannot in the group most likely to be scanned early: the great vessel origins, tandem intracranial disease, and the dissection that is a leading cause of stroke in younger patients and is invisible to a bifurcation-only examination. Two constraints belong on the request. This must not delay or displace the brain imaging that establishes the territory of the event, and the surgical question only exists in a patient who would actually be a candidate for intervention — imaging someone who would never be operated on measures something nobody will act on. |
| First line | Carotid Doppler ultrasound Carotid duplex ultrasound usually appropriate | Duplex is co-first-line here rather than displaced, and a request for it in the first 24 hours accepts as written: where a vascular laboratory can scan the same day, it grades the stenosis from the velocities with no radiation and no iodine, and adds plaque surface morphology and vertebral flow direction. What this fork says is only that it should not become the rate-limiting step — a duplex slot in three days, in a service that has already put the patient through a CT angiogram, is a delay bought for nothing. Where the two disagree, or where the bifurcation is heavily calcified or the stenosis borderline against the operative threshold, agreement between two modalities is the usual local standard before an operation. |
| Reasonable alternative | MR Angiogram Head and Neck MRA Carotid / Neck — contrast-enhanced | The radiation-free and iodine-free equivalent, and the natural choice where the patient is having brain MRI anyway — diffusion-weighted imaging confirms the territory of the event in the same sitting, which is what makes a stenosis symptomatic and therefore operable. It grades the same lesion, tends to exaggerate stenosis where flow is turbulent, and is slower to arrange in an urgent pathway, which is why it sits behind the two studies above rather than alongside them. |
- This fork is about access to a decision, not about the superiority of one modality. A service with same-day duplex has no reason to change what it does; a service without one should not let the scan slot become the reason an operable patient misses the window.
- Time since onset is also what makes the event recent enough to be worth operating on at all. Symptoms months old are a different question, with a much weaker case for intervention, and they resolve on the default pathway below.
Adults
Matches your inputsDefault| Role | Study & protocol | Why this answers the question |
|---|---|---|
| First line | Carotid Doppler ultrasound Carotid duplex ultrasound usually appropriate | Duplex infers the degree of narrowing from what the blood does rather than from what the lumen looks like: flow through a fixed narrowing must accelerate to preserve volume, so peak systolic velocity in the internal carotid rises predictably with stenosis, and the ratio to the common carotid corrects for the patient's own cardiac output. It needs no contrast and no radiation, can be done at the bedside within hours, and it also reports plaque surface morphology and the direction of vertebral flow. Its weaknesses are physical: dense shadowing calcification hides the very segment being measured, and velocity thresholds vary between laboratories, which is why the report must state the criteria used. |
| Reasonable alternative | CT Angiogram — Carotid and Vertebral (Arch to Vertex) CT Angiogram — Aortic Arch to Vertex | Direct luminal imaging from the arch to the vertex, which answers three things duplex cannot: the state of the great vessel origins, tandem intracranial disease, and the anatomy of a heavily calcified or tortuous bifurcation that shadowed on ultrasound. Caudocranial acquisition and a right arm injection are used specifically to keep dense venous contrast away from the thoracic inlet, where the origins are assessed. It is frequently already available in hyperacute stroke pathways, which in practice makes it the fastest route to a decision. |
| Reasonable alternative | MR Angiogram Head and Neck MRA Carotid / Neck — contrast-enhanced | Contrast-enhanced MR angiography grades the same stenosis without ionising radiation and without iodinated contrast, and can be combined with the diffusion-weighted brain imaging that confirms the territory of the event. It is the reasonable second modality when duplex is equivocal and iodinated contrast is a problem; its limitation is a tendency to exaggerate stenosis where flow is turbulent, and it is slower to arrange in an urgent pathway. |
- Two modalities agreeing is the usual local standard before an operation, because a single equivocal measurement is not a sufficient basis for endarterectomy.
Pitfalls
- Reporting a percentage without naming NASCET or ECST. The two methods give substantially different numbers for the same artery, and the treatment thresholds are method-specific.
- Grading a heavily calcified bifurcation on duplex where shadowing obscures the lumen, or on a thick maximum-intensity projection where calcium blooms — both systematically over-call stenosis.
- Missing near-occlusion. A trickle of distal flow with a collapsed internal carotid can produce velocities that look unimpressive, and the management is different from a conventional tight stenosis.
- Imaging the asymptomatic side and generating an operation for it. The evidence for intervention is far weaker in asymptomatic disease.
- Letting the scan slot become the rate-limiting step. Beyond about two weeks the number needed to treat rises steeply, and a slower but more elegant test is the wrong choice.
Priors — what to pull first
- Look for previous carotid imaging: a known long-standing occlusion on the symptomatic side removes the surgical question entirely, and a previously documented stenosis lets progression be judged.
- Retrieve the brain imaging from the index event. Establishing that the infarct is in the carotid territory of the stenosis is what makes the stenosis symptomatic, and only symptomatic stenosis carries the surgical benefit.
What makes a good request
- A good request states the vascular territory of the symptoms and which side, the date and time of onset, and confirms that the patient is a candidate for intervention. Imaging a patient who would never be operated on measures something nobody will act on.
- The treatment thresholds are measurement-method specific: 50 to 99 per cent by NASCET criteria, or 70 to 99 per cent by ECST criteria. A report that gives a percentage without naming the method is not actionable.
- The pathway is time-critical. National guidance places assessment and referral within a week of symptom onset with intervention within two weeks, which is why many hyperacute stroke services now acquire an arch-to-vertex CT angiogram at first presentation instead of waiting for a duplex slot — a defensible local divergence from a duplex-first policy.
Scoring this once it is done
The classification and risk tools this question ends in.
How these studies are acquired
Contrast, phases and timing for every study on the pathways above.
Confirm locally
- Carotid duplex ultrasound: timings are typical — confirm against local protocol.
- Timings, contrast volumes and rates above are typical published values. Your department's protocol, scanner and patient population decide the actual numbers.
- Safety thresholds and premedication policy follow local policy where it differs from the cited guidance.
References
- NICE NG128 — Stroke and transient ischaemic attack in over 16s: diagnosis and initial management · NICE
- Urgent need to update NICE guidelines on imaging for transient ischaemic attack (Lancet 2022) · Primary literature
Implemented from the cited published sources. Educational and workflow support only; confirm against current guidelines and local policy before clinical use.